This study proposes a randomized three group intervention to evaluate the efficacy of a smoking cessation program: Teen Freshstart, Teen FreshStart plus Buddy, compared to Usual Care among pregnant adolescents who are 12- 28 weeks gestation. The conceptual framework is based on Problem Based Theory. The intervention is a developmentally appropriate group model based on Cognitive Behavioral Theory and strengthened by concepts from Problem Behavior Theory. The smoking prevalence rate among teenage pregnant women has been estimated at 50 - 62% with 60-80% continuing to smoke throughout their pregnancies. Damaging effects of cigarette smoking on the fetus are well documented. Perinatal mortality rates are twice as high in smokers than non smokers. Smoking mothers are more likely to deliver low birthweight and premature infants with the potential for other damaging physiologic effects on the infant and mother. Because of higher risk pregnancies in adolescent populations, the teen smoker presents a double obstetrical risk. There are limited studies which focus on this risk group. Additionally, there are data to suggest that teens who engage in one problem behavior participate in other problem behaviors. This group has been shown to be influenced by peers who engage in problems behaviors. Therefore, specific aims of this study are to evaluate short and ling term effects of two smoking intervention strategies developmentally tailored to the pregnant adolescent. Two hundred and sixty two teen pregnant adolescents who smoke a minimum of 1 cigarette per day, between 14 and 17, and are 12-28 weeks gestation will be randomized to three groups which will be conducted in either a school or clinic setting. These groups include: 1.) Teen FreshStart (TFS); 2.) TFS with Buddy (TFSB) (addition of peer support); 3.) Usual Care (UC) (control). Self-report and saliva cotinine levels will be analyzed to evaluate smoking behavior at baseline and cessation after 8 sessions and 1 year post-randomization. Contingency table analysis with Chi Square statistics will be performed to test the efficacy of the proposed interventions in achieving smoking cessation at post intervention, 6 weeks postpartum and one year post randomization, as well as the efficacy of the booster session at one year post randomization. Multiple logistic regression will be used to identify predictor variables, as proposed by Problem Based Theory, from the social environment, the perceived environment, and personal factors, or response to interventions. The study will help to determine the most effective and efficient methods to achieve and maintain smoking cessation which can be implemented by school or clinic nurses for pregnant adolescents who smoke.